126 lines · plain
1.. SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause)2 3===========================4BPF_PROG_TYPE_CGROUP_SYSCTL5===========================6 7This document describes ``BPF_PROG_TYPE_CGROUP_SYSCTL`` program type that8provides cgroup-bpf hook for sysctl.9 10The hook has to be attached to a cgroup and will be called every time a11process inside that cgroup tries to read from or write to sysctl knob in proc.12 131. Attach type14**************15 16``BPF_CGROUP_SYSCTL`` attach type has to be used to attach17``BPF_PROG_TYPE_CGROUP_SYSCTL`` program to a cgroup.18 192. Context20**********21 22``BPF_PROG_TYPE_CGROUP_SYSCTL`` provides access to the following context from23BPF program::24 25 struct bpf_sysctl {26 __u32 write;27 __u32 file_pos;28 };29 30* ``write`` indicates whether sysctl value is being read (``0``) or written31 (``1``). This field is read-only.32 33* ``file_pos`` indicates file position sysctl is being accessed at, read34 or written. This field is read-write. Writing to the field sets the starting35 position in sysctl proc file ``read(2)`` will be reading from or ``write(2)``36 will be writing to. Writing zero to the field can be used e.g. to override37 whole sysctl value by ``bpf_sysctl_set_new_value()`` on ``write(2)`` even38 when it's called by user space on ``file_pos > 0``. Writing non-zero39 value to the field can be used to access part of sysctl value starting from40 specified ``file_pos``. Not all sysctl support access with ``file_pos !=41 0``, e.g. writes to numeric sysctl entries must always be at file position42 ``0``. See also ``kernel.sysctl_writes_strict`` sysctl.43 44See `linux/bpf.h`_ for more details on how context field can be accessed.45 463. Return code47**************48 49``BPF_PROG_TYPE_CGROUP_SYSCTL`` program must return one of the following50return codes:51 52* ``0`` means "reject access to sysctl";53* ``1`` means "proceed with access".54 55If program returns ``0`` user space will get ``-1`` from ``read(2)`` or56``write(2)`` and ``errno`` will be set to ``EPERM``.57 584. Helpers59**********60 61Since sysctl knob is represented by a name and a value, sysctl specific BPF62helpers focus on providing access to these properties:63 64* ``bpf_sysctl_get_name()`` to get sysctl name as it is visible in65 ``/proc/sys`` into provided by BPF program buffer;66 67* ``bpf_sysctl_get_current_value()`` to get string value currently held by68 sysctl into provided by BPF program buffer. This helper is available on both69 ``read(2)`` from and ``write(2)`` to sysctl;70 71* ``bpf_sysctl_get_new_value()`` to get new string value currently being72 written to sysctl before actual write happens. This helper can be used only73 on ``ctx->write == 1``;74 75* ``bpf_sysctl_set_new_value()`` to override new string value currently being76 written to sysctl before actual write happens. Sysctl value will be77 overridden starting from the current ``ctx->file_pos``. If the whole value78 has to be overridden BPF program can set ``file_pos`` to zero before calling79 to the helper. This helper can be used only on ``ctx->write == 1``. New80 string value set by the helper is treated and verified by kernel same way as81 an equivalent string passed by user space.82 83BPF program sees sysctl value same way as user space does in proc filesystem,84i.e. as a string. Since many sysctl values represent an integer or a vector85of integers, the following helpers can be used to get numeric value from the86string:87 88* ``bpf_strtol()`` to convert initial part of the string to long integer89 similar to user space `strtol(3)`_;90* ``bpf_strtoul()`` to convert initial part of the string to unsigned long91 integer similar to user space `strtoul(3)`_;92 93See `linux/bpf.h`_ for more details on helpers described here.94 955. Examples96***********97 98See `test_sysctl_prog.c`_ for an example of BPF program in C that access99sysctl name and value, parses string value to get vector of integers and uses100the result to make decision whether to allow or deny access to sysctl.101 1026. Notes103********104 105``BPF_PROG_TYPE_CGROUP_SYSCTL`` is intended to be used in **trusted** root106environment, for example to monitor sysctl usage or catch unreasonable values107an application, running as root in a separate cgroup, is trying to set.108 109Since `task_dfl_cgroup(current)` is called at `sys_read` / `sys_write` time it110may return results different from that at `sys_open` time, i.e. process that111opened sysctl file in proc filesystem may differ from process that is trying112to read from / write to it and two such processes may run in different113cgroups, what means ``BPF_PROG_TYPE_CGROUP_SYSCTL`` should not be used as a114security mechanism to limit sysctl usage.115 116As with any cgroup-bpf program additional care should be taken if an117application running as root in a cgroup should not be allowed to118detach/replace BPF program attached by administrator.119 120.. Links121.. _linux/bpf.h: ../../include/uapi/linux/bpf.h122.. _strtol(3): http://man7.org/linux/man-pages/man3/strtol.3p.html123.. _strtoul(3): http://man7.org/linux/man-pages/man3/strtoul.3p.html124.. _test_sysctl_prog.c:125 ../../tools/testing/selftests/bpf/progs/test_sysctl_prog.c126